Scholarlink

<p>As network traffic volumes, service diversity, and dynamism continue to surge, traditional IP/MPLS traffic engineering struggles with static metrics, per-flow signaling overhead, and constrained path flexibility. Segment Routing (SR) addresses many of these limitations by encoding explicit paths in packet headers and collapsing per-tunnel state. Yet, its NP-hard segment-list generation, hardware-imposed stack-depth caps, and evolving control-plane workflows demand adaptive solutions. This survey delivers the first systematic exploration of how artificial intelligence (AI) can bolster SR, from traffic classification and segment-list computation to fast reroute, service-function chaining, and multi-domain orchestration, by spanning supervised and unsupervised learning, reinforcement learning (value-based, policy-gradient, actor-critic, meta and federated schemes), evolutionary and swarm-intelligence heuristics, and hybrid pipelines that fuse forecasting, neural optimization, and heuristic search. We synthesize related state-of-the-art works to uncover enduring challenges and then outline concrete research directions to chart a path toward transparent, resilient, and sustainable self-optimizing networks capable of meeting tomorrow’s stringent service-level and sustainability goals.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI meets SR: a survey on enhancing segment-routing performance with artificial intelligence

  • 期刊论文

  • 发表日期：

    2026年7月27日

查看全文
  • Noha W. Hassan,
  • Mahmoud I. Khalil,
  • Hazem M. Abbas

摘要

As network traffic volumes, service diversity, and dynamism continue to surge, traditional IP/MPLS traffic engineering struggles with static metrics, per-flow signaling overhead, and constrained path flexibility. Segment Routing (SR) addresses many of these limitations by encoding explicit paths in packet headers and collapsing per-tunnel state. Yet, its NP-hard segment-list generation, hardware-imposed stack-depth caps, and evolving control-plane workflows demand adaptive solutions. This survey delivers the first systematic exploration of how artificial intelligence (AI) can bolster SR, from traffic classification and segment-list computation to fast reroute, service-function chaining, and multi-domain orchestration, by spanning supervised and unsupervised learning, reinforcement learning (value-based, policy-gradient, actor-critic, meta and federated schemes), evolutionary and swarm-intelligence heuristics, and hybrid pipelines that fuse forecasting, neural optimization, and heuristic search. We synthesize related state-of-the-art works to uncover enduring challenges and then outline concrete research directions to chart a path toward transparent, resilient, and sustainable self-optimizing networks capable of meeting tomorrow’s stringent service-level and sustainability goals.